J'accuse

[Dik T. Winter]

> In
> <455d844f-bebd-4740-a939-14a756906839@y10g2000prf.googlegroups.com>,
> spinoza1111 wrote: ....
>
> Yes, of course it does. I can write a C program without reference to
> any implementation. I can read a C program without reference to any
> implementation. I can use C to communicate programming ideas to other
> people, with no implementation being involved at all.

Indeed. Algol 60 did exist only because of that idea: communication of
algorithms to others. That it was ultimately implemented on quite a few
machines was incidental. Yes, really. When the ACM decided that
algorithms in the CACM should be written in Algol 60 there was no real
implementation around. As Esger Dijkstra (spinoza1111 likes him a lot
it seems) wrote (parafrased): computer languages are to communicate
algorithms between people.

 

[jacob navia]

Kelsey Bjarnason a écrit :

[snips]

The C standard Appendix J.5:

Common extensions

J.5.6 Other arithmetic types

1 Additional arithmetic types, such as __int128 or double double, and
their appropriate conversions are defined (6.2.5, 6.3.1). Additional
floating types may have more range or precision than long double, may be
used for evaluating expressions of other floating types, and may be used
to define float_t or double_t.

"I answered that my compiler provides"... not C, *your compiler*. Will
the code work, unmodified, in a Watcom, Borland, or Gnu compiler, invoked
in conforming mode?

No?

I repeat: he asked how to do something in *C*, not in Jacob Navia's pet
compiler.

Look, this is OK. You do not like it?

I do not give a dam about your opinion.

End of story, you are just repeating nonsense. My implementation is C.
The standard recognizes those extensions as "common extensions" and
explicitely mentions them.

Of course when gcc implements __int128 it will OK for you.

Until then, goodbye

 

[jacob navia]

Dik T. Winter a écrit :

I agree that Jacob Navia is doing something similar, the main difference
is that he asks for money for some kind of uses.

I see a tiny other difference: I do not have any well paid job in
academia, and I have two kids.

I have to survive with the work I furnish for others, as many
other people. Still I manage to give my software for free.

There have been more than a million downloads since I started.

 

[jacob navia]

Dik T. Winter a écrit :

One additional point:

Everybody is able to get and use gcc for free, even if it is for
commercial and/or educational use. That is, the compiler is actually
free and comes with complete source. And nobody from the developers
will ever earn anything for their work.

Maybe.

When I implemented a JIT for linux, I asked one of the people that
wrote that code a question.

The answer was that I needed a yearly software maintenance contract
from redhat for US$ 20 000.

This was significantly higher as Microsoft, that asks for US$ 2 000
for 3 questions in their MSDN contracts.

Eventually, I solved it without asking questions. And no, the source
code of the compiler doesn't tell you anything.

There isn't a single comment or technical description.

It is the generated code + hours in the debugger that led me to
the solution.

No, I am not telling you that what that developer did was wrong,
since they have to live from their work as I do. What I am telling
you is that it is the same thing I do.

When you mention Redhat, Suse and others: what they provide (and what
is not so easy to obtain, but can be done), is a complete operating
system.

They did not paid for the development of that system. The biggest part
was done by unpaid programmers.

They provide you with it, complete, and they ask money for
that. However, if you are willing to obtain the individual parts and
glue them together, you can do so for free, which I have done for early
versions of gcc and other gnu-tools.

Sure, but it will cost you so much work that nobody will do it.
Their reasoning is sound. That is why there are no comments in
GNU source code in the majority of applications. No comments, no
description, and it will change in a whim.

Why?

Think for a moment.

My institute elected to supply some 200 of the about 250 computers
present with variants of Linux, all from bundled versions, like Fedora,
Debian, Suse and some others. (The remainder, mostly not used in
research have some version of Windows installed.)

Your institue paid for commercial use then. Why is so bad that I do the
same?

 

[Chris M. Thomasson]

Dik T. Winter a écrit :


I see a tiny other difference: I do not have any well paid job in
academia, and I have two kids.

I have to survive with the work I furnish for others, as many
other people. Still I manage to give my software for free.

I hope you make a shi% load of cash from you're hard works Sir!

:^)

 

[Nick Keighley]

Nick Keighley a écrit :

if it isn't open source (I'm not certain of this but I couldn't find
a source download on the site either) in what sense is it "open".

?

and which variant of "free" are you using. If I can't is it "free"
in the Stallman sense?

[note: I don't care if it's "open", "free" or whatever. Jacob can do
as he please with his own stuff. I'd just like it to be described
correctly]

did you read that bit. That was put in precisely to try and stop you
throwing a hissy-fit.

Contrary to redhat, suse, and other companies that sell the work
of others, I sell only my work.

good for you. I hope you make some money.

My software is distributed for free since 12-13 years.

FOR NON-COMMERCIAL USE.
Why do you always miss that bit out?

I have financed
all expenses, put up with all the work. I refuse however to give it away
as GNU source so that redhat or another company that thrives on open
source can sell it and profit from my work with me receiving NOTHING.

good for you. Did you see the bit where I said "I don't care if it's
open
source"?

Because here there is obviously a double standard.

not on my part. If people want to make money from their labours
this is good. If people want to give stuff away this too is good.
You stuff, do what you want with it.

<snip more similar crap>

 

[Nick Keighley]

Kelsey Bjarnason a écrit :


The C standard Appendix J.5:

Common extensions

the bit you missed [I quote from the ANSI 89 standard, I assume
similar
wording in C99]

"The following extensions are widely used in many systems, but are
not
portable to all implementaions. The inclusion of any extension may
cause a
strictly conforming program to become invalid [is there a word missing
here?]
renders an implementation non-conforming. [...]"

[errors in transcription are mine]

J.5.6 Other arithmetic types

1 Additional arithmetic types, such as __int128 or double double, and
their appropriate conversions are defined (6.2.5, 6.3.1). Additional
floating types may have more range or precision than long double, may be
used for evaluating expressions of other floating types, and may be used
to define float_t or double_t.

Before spewing nonsense turn on your brain, read the standard, think
a bit.

you should take your own advice

 

[Chris M. Thomasson]

Dik T. Winter a écrit :


I see a tiny other difference: I do not have any well paid job in
academia, and I have two kids.

I have to survive with the work I furnish for others, as many
other people.



Still I manage to give my software for free.

Humm... I wonder what would happen if somebody somehow acquired you're
compiler, and used it commercially but did not pay you anything because
she/he read that the compiler creator explicitly stated that his compiler
was free. Would they be in error in any way, shape or form Jacob?

 

[jacob navia]

Chris M. Thomasson a écrit :

Humm... I wonder what would happen if somebody somehow acquired you're
compiler, and used it commercially but did not pay you anything because
she/he read that the compiler creator explicitly stated that his
compiler was free. Would they be in error in any way, shape or form Jacob?

That has happened countless times. For instance, the University of Nice
had a copy of my software without paying me a single penny.

When I protested they did not cared to even answer my protests.

Other universities have been using it too, since I know when the
students ask me questions about it, or even want that I do their
homework!



Many schools ask me for authorization since with disminishing
budgets they can't afford licenses. Sometimes I agree, sometimes
not.

Of course if somebody is making a lot of money with my software
I would sue him/her.

 

[Dik T. Winter]

> Dik T. Winter a écrit :
>
> When I implemented a JIT for linux, I asked one of the people that
> wrote that code a question.

About what code are you talking? Is the code you did ask about free and
open source?

>
> They did not paid for the development of that system. The biggest part
> was done by unpaid programmers.

Right, they provide additional service, and if you do not want that
additional service, you can do without it.

>
> Your institue paid for commercial use then. Why is so bad that I do the
> same?

That is not bad. It is when you state that your compiler is free that you
are wrong. It is only free in a limited number of circumstances. And
compare that to gcc which actually *is* free for everybody to obtain for
whatever purpose.

 

[spinoza1111]

Nick Keighley a écrit :

On Sep 9, 6:04 pm, Nick Keighley <[email protected]>
wrote:
Like many creators of open source or "free" software, Navia is
perfectly willing to let people use his software noncommercially for
free, but would like businesses to pay. This is not "hypocrisy". How
dare you, sir, make this charge.
is lcc-win32 open source? It certainly ain't "free"
http://www.cs.virginia.edu/~lcc-win32/(Navia'sterms)http://opensource....)
Navia is what we'd commonly understand as open software, but it is not
open source,
if it isn't open source (I'm not certain of this but I couldn't find
a source download on the site either) in what sense is it "open".

?

since there's no indication at the above site. He may
give us more correct information in this discussion.
and which variant of "free" are you using. If I can't is it "free"
in the Stallman sense?
[note: I don't care if it's "open", "free" or whatever. Jacob can do
as he please with his own stuff. I'd just like it to be described
correctly]

did you read that bit. That was put in precisely to try and stop you
throwing a hissy-fit.

If defending yourself has become "throwing a hissy fit", then male and
female have reversed meaning. The expected behavior here is to fawn on
the self-appointed experts. Navia doesn't.

good for you. I hope you make some money.


FOR NON-COMMERCIAL USE.
Why do you always miss that bit out?

If his software is distributed free to some parties, then it is true
that his software is distributed for free. It is probably untrue that
his software is free, but he doesn't make that claim, although
arguably IF software is free to enough people, then it is free
software.

Here, the deficient reading comprehension of his enemies is
instinctively used by them in a perverse way. As in the case of the
obscene spectacles of last summer's "town meetings" in the USA, people
don't know that they're uninformed or deficient in comprehension and
instead these very deficiencies become a sort of brute strength. Here,
you use a willful failure to comprehend to call Navia a liar and you
confuse brutality with precision.

 

[spinoza1111]

No.  Unacknowledged self-promotion is problematic, but for reasons
other than topicality.

You have been indoctrinated by the needs of corporate employment to
treat Navia's syntax (in which Navia is always, of necessity and
justifiably, using the syntax "I did x", putting himself in the
driver's seat and asking, not so much for money as for recognition of
this fact.)

But in the American-style corporation, people must efface themselves
lest they have ownership claims to their intellectual product, as
witness the extraordinary cruelty of "nondisclosure" agreements that
are forced upon many programmers as a condition of employment, in
which the programmer may not even develop software of his own
invention on his own equipment and on his own time.

So instead of allowing Navia his fifteen minutes of fame in
tranquility, your instinct, you having been cheated yourself, is to
take him apart...and to call people "trolls".

 

[spinoza1111]

 > Dik T. Winter a écrit :
 > > Everybody is able to get and use gcc for free, even if it is for
 > > commercial and/or educational use.  That is, the compiler is actually
 > > free and comes with complete source.  And nobody from the developers
 > > will ever earn anything for their work.
 >
 > When I implemented a JIT for linux, I asked one of the people that
 > wrote that code a question.

About what code are you talking?  Is the code you did ask about free and
open source?

 > > When you mention Redhat, Suse and others: what they provide (and what
 > > is not so easy to obtain, but can be done), is a complete operating
 > > system.  
 >
 > They did not paid for the development of that system. The biggest part
 > was done by unpaid programmers.

Right, they provide additional service, and if you do not want that
additional service, you can do without it.

 > > My institute elected to supply some 200 of the about 250 computers
 > > present with variants of Linux, all from bundled versions, like Fedora,
 > > Debian, Suse and some others.  (The remainder, mostly not used in
 > > research have some version of Windows installed.)
 >
 > Your institue paid for commercial use then. Why is so bad that I do the
 > same?

That is not bad.  It is when you state that your compiler is free that you
are wrong.  It is only free in a limited number of circumstances.  And
compare that to gcc which actually *is* free for everybody to obtain for
whatever purpose.

This fails to correspond with common usage. Nobody expects a
corporation to get an "enterprise edition" for free. However, in
contrast with, say, Flight Simulator, Navia's software is free.

As technicians, you folks collectively seem to have a primitive,
almost theological understanding of the relationship of names to
reality, but the adjective "free" gets its meaning from the difference
between Jacob's software and Flight Simulator.

 

[spinoza1111]

It isn't bad that you charge for your software. It's bad that you tout

Bullshit. Unlike you and your friend Keith, he accomplished something
and is sharing it for free in a technical newsgroup where his product
is of high interest. We want to know about this. We want to use the
basics of his compiler, which certainly compiles recognizable C, and
we want to know how it might address C's shortcomings, where C's
shortcomings are on-topic, unless this is the Chinese politburo.

These newsgroups are a perfectly appropriate place to announce free
software, including software with a free base release and plus-cost
extensions (which is true of most Open Source).

You attack Navia because he's done something you haven't and because
like a man he defends himself, instead of snipping at others like a
little baby girl.

your product in a technical newsgroup, and it's bad that you
misleadingly claim it's "free" when in fact you charge for it in some
cases.

Bullshit. BULLSHIT. If the base release is free, it is free software
with a plus-cost extension.

Again, the fact that you charge is not a problem. Everybody
charges for their work. But we don't get P J Plauger in here pushing
his library.

That's because most eminent computer scientists, with the significant
exception of John McCarthy, are afraid of the vicious and vile
comments you losers like to make, like that of US Representative Joe
Wilson during Obama's speech today.

We don't get Thomas Plum in here pushing his validation
suite. We don't get Walter Banks in here pushing his C compilers. And
none of them posts misleading information in here about charging

You are lying. Navia has always been perfectly open about what's free
and what's not.

structure. Note: they all charge for their product - so that is
clearly NOT the issue. The issue is two-fold (a) misleading
information about charging; (b) inappropriate use of a technical
newsgroup to promote your product. You have been told this many
times.

Go to hell.

 

[Kenny McCormack]

....
(some self-important doofus wrote)
It is amazing how often this nitwit (and his spiritual comrades) says
this phrase "You've been told this many times" (or variants, such as
"I've explained this many times"), as if it had any intellectual
currency at all. I.e., as if either of the following carry any weight
at all:
1) That the speaker (i.e., Dicky, Kiko, etc) said it.
2) That they have said it multiple times (as if that saying
something multiple times somehow magically makes something true).

It is also amazing and sad that this approach works on so many people.
So many people accept the above ridiculous assertions as true.

Go to hell.

Indeed. This is about the only sensible reaction to their posts.
They need to be told this more often.

 

[Tim Streater]

spinoza1111 <[email protected]> wrote a lot of good stuff,

No, he wrote twaddle as usual. We've been pestered with his cobblers in
other NGs. He's just a nasty bullying fascist.

but I'm
only commenting on one particular thing:
...
(some self-important doofus wrote)

It is amazing how often this nitwit (and his spiritual comrades) says
this phrase "You've been told this many times" (or variants, such as
"I've explained this many times"), as if it had any intellectual
currency at all.

The "this" in question will have been explained or pointed out
innumerable times without refutation or sensible counter-argument.

 

[Keith Thompson]

Kelsey Bjarnason a écrit :

Topicality is essential to Usenet's users. I know that you
don't understand this. Destruction of personalities is not
topical here, so please do it in a newsgroup where it is
topical.

My answer was on topic. The user asked how he could use 128 bit
numbers in C. I answered that my compiler provides native
support for 128 bit numbers.

And, of course, your compiler defines C, right? After all, the
question was how to do it _in C_, not in _Jacob Navia's pet
compiler_. Unless the two are synonymous, your answer would have
been wholly irrelevant.

The C standard Appendix J.5:

Common extensions

the bit you missed [I quote from the ANSI 89 standard, I assume
similar
wording in C99]
[snip]
[errors in transcription are mine]

My copy reads: "The following extensions are widely used in many
systems, but are not portable to all implementations. The inclusion
of any extension that may cause a strictly conforming program to
become invalid renders an implementation nonconforming. Examples of
such extensions are new keywords, or library functions declared in
standard headers or predefined macros with names that do not begin
with an underscore."

[...]

I've been curious about just how lcc-win implements 128-bit integers,
particularly whether they're implemented as an extension that doesn't
violate the C99 standard. I don't currently have a system on which
I can install it, so I resorted to Google and found the following:

In <http://www.q-software-solutions.de/products/lcc-win32/changelog.shtml>:

128 bit integers are now possible. To use them #include <i128.h>.
The library is i128.lib.

(Incidentally, I had to view the HTML source to read that; the '<' and
'>' characters need to be converted to be read properly in
an ordinary browser window.)

And at
<http://coding.derkeiler.com/Archive/C_CPP/comp.lang.c/2008-04/msg04961.html>,
apparently somebody's archive of his newsgroup, in an article posted
in April 2008:

P.S. This feature is now native in the 64 bit version, i.e.
not using operator overloading as in the 32 bit version.

The good news is that if the feature isn't enabled until you add
"#include <i128.h>", then it doesn't break any strictly conforming
code.

The not quite as good news is that, at least in the 32-bit version
of lcc-win, it's apparently not implemented using "extended integer
types" as described in C99 6.2.5.

I would like to encourage *some* implementer to provide
128-bit integers using this mechanism. This would mean that no
system-specific header would be required, that intmax_t and uintmax_t
would be 128-bit types, also defined as {,u}int128{,fast,least}_t,
and that *portable* code could potentially use 128-bit arithmetic.

In general, I don't know of *any* C compilers that provide extended
integer types as defined in C99 6.2.5 -- but I'm certainly not
familiar with all C compilers.

This is not (necessarily) a criticism of the way lcc-win currently
does this. Providing it as a non-portable extension is probably
a good start.

(Possibly the 64-bit version of lcc-win already does this, but I
haven't found enough information to confirm that.)

 

[jacob navia]

Richard Harter a écrit :

One can take the view, and I do, that there is no single language
called C. Rather there is a suite of implementations clustered
around several focal points, e.g., the two ISO standards, K&R I,
and whatever Microsoft recognizes as C. In my view, it is
perfectly legitimate for Jacob to say that he has an
implementation of C, and that it provides a variant of C. It
isn't a conforming implementation of either ISO standard, but I
don't think Jacob is claiming that. (I may be wrong.)

As a side note, when one takes into account bugs, it may be that
there are no comforming implementations.

I try to implement full C99, but I am not there yet, each day
having only 24 hours...

I have implemented several features lately, that were completely missing.

But I have problems still with unnamed initializations of immediate
data in some contexts.

struct foo {int x;int y;};
struct foo baz = (struct foo){1,2};

FP_CONTRACT is missing, I have some problems in the header files, etc.

 

[Richard Bos]

gwowen a écrit :

I bought the copyrights from Addison Wesley for an important sum.

That may make it your copyright, but it doesn't make it your work.

Thisof course doesn't mean that I think their work is bad or
unimportant.

No, but you do make a point of never mentioning them until called on
your dishonesty.

Richard

 

[jacob navia]

Richard Bos a écrit :

That may make it your copyright, but it doesn't make it your work.


No, but you do make a point of never mentioning them until called on
your dishonesty.

Richard

This is blatantly not true. In the home page of lcc-win they are
mentioned in the second sentence.

I quote

lcc-win32: A Compiler system for windows
by Jacob Navia
based on the original screenplay of
Dave Hanson and Chris Fraser: A portable C compiler